Texas power issues. Windfarms getting iced up. 67

[Alistair_Heaton]

That's a lot of capacity that's been lost. Must admit being Aberdonian we always associate Texas with hot and humid... Not snow storms.

https://www.independent.co.uk/news/...nd-turbines-texas-power-outages-b1802596.html

 

[waross]

Hydro plants and water hammer.
The Russian plant had a timed closing that had been disabled or bypassed for some reason.
I have seen designs for Pelton wheels where the first action of the governor was to swing a diverter in front of the nozzle to divert some flow away from the bucket wheel. As the valve was slowly closed, the diverter would retract.
Some plants use surge towers and surge tanks to dissipate the kinetic energy of the moving water column.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[waross]

I come from a time when cars used to have 6V batteries...

Remember when they invented turn signals and brake lights?
When I took my first driver's test, we had to use hand signals when turning.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[JohnRBaker]

I can't remember cars without brake lights, but the fist car I ever attempted to drive, you had to press the starter button with your left foot. And the radio was AM only and there were no push-buttons.

John R. Baker, P.E. (ret)
EX-Product 'Evangelist'
Irvine, CA
Siemens PLM:
UG/NX Museum:

The secret of life is not finding someone to live with
It's finding someone you can't live without

 

[bacon4life]

What they're saying is if the operators did not take prompt action to reduce load by initiating partial blackouts, then the entire Texas grid could have gone down, tripping every generating plant offline.

Yes, the grid operator had to shed load to keep the grid running. The ERCOT Underfrequency Load Shedding Plan only requires 25% of firm load (plus some non-firm loads) to be armed for UFLS, which is far less than the 49% of generation lost during the event. UFLS is primarily designed to keep generation turbines from experiencing damage from mechanical resonances. From the ERCOT manual :

"ERCOT provides settings for under-frequency tripping (if installed) of steam turbine generators.
No tripping should occur as long as frequency remains above 59.4 HZ. As frequency falls
below 59.4 HZ, progressively faster time delayed tripping can occur. Once frequency reaches
57.5 HZ or below, the generator can be tripped with no intentional time delay."

After a generator outage, the system operator must quickly return the frequency to 60 Hz in order to ensure the frequency stays above the generator under frequency tripping set points for the NEXT outage. If the system were allowed to persist operating at a lower operation point a small additional frequency dip could lead to generators tripping. Today's ERCOT presentation states that the system was below 59.4 Hz for more than 4 minutes and dropped as low as 59.3 Hz.

A few things make it important to keep large safety margin regarding UFLS:
1) The load armed for UFLS may not be representative of the overall load mix. Having a large amount of irrigation load armed for a summer UFLS plan is worthless during the winter. Likewise, arming heating loads for a winter UFLS plan will underperform during a summer outage.
2) UFLS rarely operates, so there is little opportunity to compare actual performance to the predicted performance. Last time it operated within my region, there were many lessons learned.
3) The off nominal frequency behavior of both generation and load is not precisely modeled in power grid simulation software. Linearizing assumptions often help simplify the models for a small ranges of frequency but at the cost of inaccuracy for large frequency variations.
4) The inertia of the grid changes as the types of generation shifts. This occurs both on an hourly basis throughout the day and due to long term trends.
5) The inertia of the grid continues to be reduced as more motors are moved variable frequency drives (VFDs).
6)Loads armed for UFLS might also be tripped for other reasons. It is challenging to avoid double counting UFLS loads when the same load might also be part of Under Frequency Load Shedding, have a manual rotating load shed plan, and be experiencing a weather related outage.

As to how long a black start might take, today's comment from ERCOT’s CEO seems pretty spot on to me:

"If we have a blackout of the system, the system is out for an indeterminate amount of time, and it's extraordinarily difficult to bring it back. We might still [5 days later] be talking about when's the power coming back on if we let the system get to that condition."

 

[byrdj]

the "Nice" size steam turbines of the 80's design were designed assuming grid frequency would be fairy stable.

Off frequency operation would place the longer buckets (blades to layman or if reaction design) into resonace vibration and just a few minutes operation would make breakage a high probibility. Part of the turbine monitoring were totalizers to recorded accumlate time at off frequency.

I recall having to discuss this when operations observed those requirements to "hang on" durring upsets.

also, ramping a steam turbine to compensated for system load swings expend the cyclic life of the rotating componets (ie, bending the paper clip) thus the more ramps, the more maintence replacing components (or failures)

 

[1503-44]

ERCOT officials stated that natural gas plants failed the most in the cold weather. Wind generators also had trouble, but overperformed at times. "There were a lot of issues around gas supply during this event, Magness said. "What I like to emphasize here is the storm affected every generation type."

 

[itsmoked]

I know our town stores NG in a serpentine pipeline in the PG&E service yard. I think it's good for running town for something like 8 hours. That's stored at high PSI.

Can you store NG in those floating tanks for a power plant? Or is that unrealistic with maybe the plant consuming a tank's worth in a day? I'm sure those tanks store it at a hideously low pressure.



Keith Cress
kcress -

http://www.flaminsystems.com

 

[LittleInch]

BI - Of course I would. I've always found it interesting that an ESD is something that is often designed and never tested in anger because of the potential consequences....

But at least you didn't overtop the relief tank eh?

If you get into column separation and vacuum collapse then you can see massive spikes and flow disturbance going back and forward. I guess sitting there and just watching the pressure climb and climb with no way of stopping it is kind of scary.

A bit like the grid operators we're talking about. Then you need to make really fast decisions or just let the automatic systems do their job and then deal with the consequences.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Gases and liquefied gases of any pressure must be stored in spheres or bullet tanks, although they are at times stored at near atmos pressure in "gas holder" tanks.

https://en.m.wikipedia.org/wiki/Gas_holder

Floating or fixed roof tanks are very low pressure, basically just for containment of the liquids themselves and fumes from slight off gassing or to control evaporation losses.

LI, .. Looking over your shoulder for places to take cover.
They must call them relief valves, because you are very relieved when they finally stop.

 

[LittleInch]

Keith,

It must be a might small town if they can build up enough gas to feed it for 8 hours.

Or a very big serpentine.

The big gas holders which go up and down are at only a few inches water columns (9 or 12 inches from memory)

So a) they would need pressuring to about 25 or 30 bar for injection into the gas turbines and would last about 5 minutes max.

Gaseous gas storage is about as difficult as storing electricity. Can be done but very expensive and only really designed for peak shaving, not wholescale replacement and wouldn't last long if it did (hours rather than days).

Some European countries have significant gas storage as strategic reserve as the rely on Russia to be friendly and not turn off their supply, but even then it's expensive to do and will only last a few days / weeks maximum.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[JJPellin]

We store liquid butane in a cylindrical, fixed roof tank. The tank holds 500,000 barrels. We keep the butane liquid and keep the pressure low by refrigerating the tank. The refrigeration systems is quite large and complex. We use the butane as a gasoline blending stock in winter. If we wanted to vaporize it to burn in our heaters, it would require another system for vaporization.

This tank is gradually filled over the course of the summer and then blended into gasoline over the course of the winter. So, it is always being filled or drained, depending on the season.

Before we built this tank, we used to ship our butane to Texas by pipeline where they pumped it into deep underground salt domes. So, Texas has the capacity to store large quantities of LPG.

Johnny Pellin

 

[davefitz]

Most large frame gas turbines require inlet fuel gas pressures between 500-900 psig ( 34-61 atm) at the inlet to their fuel gas control system. A base loaded designed plant may be provided with a fuel gas booster compressor to adress low pipeline gas pressure, but many of the simple cycle "peakers" might not have such booster compressors, as the normal interstate pipeline presure may be over 900 psig. In the case of this recent Texas fiasco, the gas pipeline pressure dropped , likely below the minimum pressure needed to startup or operate the "peakers", which led to a cascade of loss of peaker generation. Either a backup liquid fuel should be stored at those sites or the gas pipeline freeze up problems need to be adddressed.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick

 

[LittleInch]

The issue usually isn't storage, but the capacity to turn this into (very) large quantities of gas very quickly.

Same thing as all the other issues - No one wants to pay for equipment which is used for 10 days every 5 years.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[cranky108]

Are there any other gas storage/replacement systems in operation, other than the Air/Propane plant my company operates?
Now because of the rate caps, our company is discussing how to recoop the high gas prices we had to pay for electric generation. Maybe we should have used coal, as that price is fixed.

As above, water hammer is an issue at hydro plants, and deverter vanes are used for fast tripping of a newer hydro unit. The older units used a second lower valve that would open, and miss the buckets.

 

[1503-44]

cranky, did you have an issue with loss of gas supply, gas prices, or both?
The solution may depend on how much gas you use and when and for what ultimate purpose. Such as building HVAC, self power gen, industrial furnace, growing tomatoes, etc. and your location relative to pipelines, rail, harbors.

 

[cranky108]

I'm not in Texas, but yes gas prices for pipeline gas did go up quite a bit.

I don't know how much distribution gas we used, or even how much gas we used at out power plants.
In fact the high gas and energy usage thing just came out early this week. I do know we did set a new Winter peak. Which usually happens around late December.

As I said, maybe we should have used coal, but I don't know at what point each of our plants are at.
Just that several of the hydro plants are frozen at this time of the year.

I honestly don't know what our customers do with the energy. It is not good business to dig into that for me.
I believe about 50% or more of the energy used are homes.

 

[dik]

Really bad idea... too big a carbon footprint.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[electricpete]

Interesting challenges about gas storage and retrieval outlined by LittleInch. I always heard it generally had to come through the pipelines at the time it was needed but never gave much thought as to why.

Different subject...
Texas’ Power Grid Was 4 Minutes And 37 Seconds Away From Collapsing. Here’s How It Happened

It's an interesting story with graph showing frequency going below 59.4hz for 4 minutes (and timeline annotated on the graph). There are repeated references to ERCOT having to tell (order) power providers to reduce the demand. That makes sense in view of bacon4life's explanation of ERCOT underfrequency load shedding where only 25% of load is equipped for automatic UF load shedding, but a higher fraction of generation tripped, so further manual actions for load shedding were required.

So avoiding that total collapse scenario relied on humans communicating/acting fast enough. Even though there was almost a 9 GW decrease in wind generation over 24 hour period, it was very gradual so probably would not contribute to that particular scenario. On the other hand if we had a large block of generation trip suddenly when we were already down at 59.4hz, who knows what would have happened (and when you look at how often there were discrete losses of generation on that timeline, it seems like just a matter of luck that no generation loss occurred during the 4 minute period below 59.4hz).


=====================================
(2B)+(2B)' ?

 

[electricpete]

A guy I work with has some buddies in the gas industry. Here is the story about why (at least some of) the gas went down... as they told him, which he told me, which I'm now telling you (3rd hand... take it with suitable grains of salt )

Apparently the gas comes out of the well with moisture in it. So they inject some kind of solution into the gas to keep it from freezing (I think of it as anti-freeze). Of course they need pumps to inject that anti-freeze solution at the proper pressure. They don't have any electricity handy, so how do they power them? They use something similar to a compressed air motor, except since they don't have compressed air, these anti-freeze injection pumps are powered by the pressure of the available gas.

BUT the gas that is used to power the gas-powered anti-freeze injection pump itself froze... because there was no provision for adding anti-freeze to the gas that powers the anti-freeze injection pump! (now I'm picturing that the pipeline tap for the gas to power the injection pump must have been upstream of the injection point rather than downstream, although it wasn't described to me in those terms).

It seems like a dumb design in retrospect, but had been that way forever and the folly was never revealed... until now.

=====================================
(2B)+(2B)' ?

 

[electricpete]

Someone described the south texas project unit trip above. I heard it was due to a 6' length of uninsulated fluid sensing line for sgfp suction pressure. That length of sensing line is hidden in a very obscure out-of-the way place that you have to squeeze to get to. Presumably the difficulty in getting to that particular length of pipe is why the insulators never finished their job. It is also easy to see why no-one ever noticed it.

=====================================
(2B)+(2B)' ?